Adam Mosseri, Head of Instagram, said the AI-powered warning feature has stopped some people from posting foulmouthed comments during the early testing period: Often, people use different tricks like using symbols or alternative spellings to fool the AI and post abusive comments. Instagram hasn’t shared any details as to what it’s doing to curb that, and hasn’t specified if the feature is available for languages other than English. We’ve asked the company for more information, and we’ll update the post accordingly. The social network is also testing another feature called ‘Restrict,’ which will allow you to limit a person’s interaction with you. If you restrict a person, they will still be able to post comments on your posts, but they won’t be visible to anyone but themselves. You can then review and allow the restricted person’s comment to be visible to others specifically. Instagram said often people don’t block, unfollow, or report their bullies, because “it could escalate the situation.” The restricted person won’t be able to see when you’re active on the platform, or when you’ve read their direct messages. In April, the platform started demoting offensive posts as a measure to curb hate speech. We’ll have to wait and see if these features are effective, and if the social network’s AI is strong enough to detect tricky abusive comments.
